Defect image datasets for industrial inspection.
What it does
Creates and labels industrial inspection image datasets by orchestrating rendering, augmentation, training, and inference workflows on OSMO.
When to use it
Use for defect image generation for PCBA, metal surfaces, or glass, including initial asset setup, finetuning, and inference on real images.
How to use it
Give it an inspection dataset request; it asks for storage and workflow choices, then runs workflows that publish results to OSMO-backed storage.
Uses
Good to know
- cannot be undone: /workspace/paidf-anomalygen/checkpoints
- edits its own instructions
What you need · 4
Requires the OSMO CLI on PATH to inspect configuration and submit workflows.
Requires Hugging Face access with the relevant gated model licenses accepted so the token can read the workflow's repositories.
Access · 6
This skill
/workspace/paidf-anomalygen/checkpoints
Replaces container checkpoint entries (irreversible)
OSMO
Starts image generation workflows
Stores credentials and generated images